Thursday, December 10, 2015

SELF-FULFILLING STOCK FORECASTS

FROM BUSINESS WORLD OF 28 SEPTEMBER 2006 


Unsolicited advice


Those who invest through brokerage houses are familiar with the recommendations they receive; so are those who read the financial press. How often do they see recommendations to buy? And how often to sell? My own impression is that most recommendations are to buy. But no money is made unless what is bought is sold; and if one waits too long, the value of one’s investment may depreciate either because the market goes down or the company loses out on account of poor management. One may also well ask: why do people give so many more recommendations to buy than to sell? Might they be agents or partners or beneficiaries in some way of the companies whose shares they are recommending? Might there be an improper nexus between them and the companies?
These questions remained unanswered in my mind until I came across a recent book written by Rajesh Chakrabarti (The Financial Sector in India: Emerging Issues, Oxford). There he takes 2190 recommendations on 310 companies by analysts from 25 firms and a newspaper between January 1998 and July 2003. Of them, 46 per cent were strong buy, 21 per cent weak buy and 4 per cent neutral/buy; thus, three-quarters were buy recommendations of one sort or another. Only 9 per cent were sell recommendations. One cannot imagine stronger confirmation of my suspicion that analysts give biased recommendations. Chakraborti does not use such a strong word; he just calls it broker optimism.
Who were the worst optimists? A number gave only a handful of recommendations. Amongst those who gave over a hundred recommendations, Motilal Oswal was the most optimistic, with 89 buy recommendations out of 106. Pioneer (83/107), LKP (158/191) and SMIFS (130/164) were not far behind.
Who were the least optimistic? Business Line (136/283) was the leader. But it was not a pessimist. Rather, it enjoyed sitting on the fence; 108 of its recommendations were neutral. Rooshnil was similar; although 24 of its 54 recommendations were strong buy, another 18 were neutral. HDFC Securities got a low score for optimism: although 54 of its 82 recommendations were to buy, they were all weak buy. Although 158 of LKP Securities’ recommendations were to buy, they also made 25 recommendations to sell. Moneypore had the largest proportion of recommendations to sell – 19 out of 125.
But maybe the optimists were right in their optimism? Chakraborti shows that they were on the average. Over the 80 days after strong buy recommendations, the stocks outperformed sensex. Not much; the rise in their prices at the end of 80 days was about 3 per cent, against 1½ per cent in sensex. But the difference was statistically significant. Similarly, over the 80 days after strong sell recommendations, the fall in the stocks was greater than in sensex.  Sensex fell about 3 per cent, and the stocks about 4 per cent; but again, the difference was significant. It was not statistically significant over the entire period of 80 days. It was significant on most of the days in the case of strong sell recommendations, but not in the case of the rest. It may be noted that sensex rose on the average after buy recommendations and fell after sell recommendations. In other words, the recommendations were partly based on a forecast of the market trend.
Did the analysts lead investors to buy or unload shares and thereby fulfil the analysts’ forecasts? To test whether they did, Chakraborti compared the average price in the five days preceding the forecast to it in the five days after the forecast, and similarly for 20 days, and found a definite correspondence between the direction of the forecast and of the price change – more in the case of sell than of buy recommendations. Thus, the herd behaviour of investors – of those investors who follow analysts – helps the latter prove right.

If analysts can influence prices, promoters and other interested parties may use them for that purpose. Do they? One could answer this question by looking at prices over a longer period than 80 days – say, a year or two. Chakraborty does not apply this test. But the lesson of his results must be that it is not a bad idea to follow the recommendations of your favourite stock astrologer, especially the strong buys or sells. But act on them immediately. And do keep an eye on analysts. Even if they are all wrong, the market will prove them right, at least in the short run.